Speech Visualization for Learning Pronunciation Based on Auditory Characteristics

Efforts to date have been devoted to speech analysis and visualization. In this study a new speech visualization method based on auditory characteristics and Fourier Transform is introduced. First, we described the biological Fourier-Based speech analysis; Then, speech visualization application based on these ideas are described. Formant difference plot of Chinese vowels or words is shown. The plot axis is the difference of the first and second formant and the difference of the third and second formant, respectively. The results show that the new plot reflects the acoustic characteristics correctly. This may be a kind way of learning pronunciation for deaf persons by visual.